Materials Science Forum Vol. 1182

Abstract: High-performance thermoplastic polymers paved the way for new fast manufacturing pro-cesses, including welding. In order to obtain optimal bonding of the substrates, an adhesion step isrequired, governed by two main phenomena : intimate contact and healing. While healing has beenvastly explored, theorized and starts to be understood, prediction and characterization of the degree ofintimate contact is still a challenge. After a review of squeeze flow models for intimate contact, alongwith the expressions of the analytical solutions for a Newtonian and a shear-thinning fluid modeled bypower law, a finite element model is presented in order to observe the influence of asperity geometry,fluid behavior, and other assumptions on the evolution of the degree of intimate contact.

Abstract: Liquid Composite Molding (LCM) processes, used for producing high-quality, complex composite parts, rely on the uniform infiltration of liquid resin into fibrous fabrics. These fabrics possess a dual-scale structure: highly porous inter-tow spaces surrounding denser fiber tows. The resulting disparity in permeability and flow rates is a primary cause of defect formation, such as voids. To minimize these defects, accurate simulation, incorporating the critical influence of capillary pressure on resin infiltration within the fiber tows, is essential. This work presents a robust numerical model developed to simulate the two-phase resin flow and impregnation dynamics within a digitized, real plain-weave E-glass reinforcement obtained via X-ray micro-computed tomography (CT). The simulation utilizes the open-source multiscale multiphase solver, hybridPorousInterFoam, which employs a Darcy-Brinkman approach, transitioning between Darcy's law in porous regions and Navier-Stokes in free space. A key methodological enhancement involved modifying the advection algorithm using the isoAdvector scheme to mitigate numerical instabilities caused by the high viscosity ratio between the resin and air. Capillary effects at the mesoscale are incorporated through multiscale parameters, specifically the drag and surface tension forces. The key findings demonstrate that the modified solver successfully handles the fluid-fluid interface advection for high viscosity ratios. A parametric study highlighted the significant effect of capillary pressure on multiphase flow within the dual-scale porous media. The numerical results for flow front advancement showed very good agreement when compared against dedicated experimental validation data, confirming the model's high predictive accuracy and its potential for optimizing LCM injection conditions.

Abstract: Residual internal stresses arise during thermal processing of thermoplastic composites due to differential shrinkage of stacked orthotropic plies and can lead to defects such as shape distortion, microcracking or delamination. In the current study, a comprehensive thermomechanical model for predicting residual stresses and strains in semicrystalline thermoplastic composites is presented with specific application to unidirectional Carbon fiber reinforced LM-PAEK composite laminates. The model is based on an incremental Classical Laminate Theory (CLT) framework that incorporates temperature-dependent material properties and accounts for both thermal and crystallization-induced shrinkage effects. Material characterization is performed to measure key temperature-dependent properties: thermomechanical analysis (TMA) is used to measure the transverse thermal expansion coefficient (CTE) and crystallization shrinkage upon cooling from melt state, dynamic mechanical analysis (DMA) to obtain the transverse modulus. The stress-free temperature, the temperature at which residual stresses begin to develop, is identified through curvature evolution measurements of unsymmetric laminates using image analysis. The model validation is performed via curvature measurements of unsymmetric cross-ply laminates using laser scanning techniques, demonstrating good agreement between model predictions and experimental measurements.

Abstract: This work primarily focuses on the development and simulation of the Liquid Resin Infusion (LRI) process for particle-filled resins, aiming to impart additional functionalities to composite parts. The paper presents both the simulation development and the experimental tests used to establish physics-based models. The main challenge lies in understanding how particle addition affects the resin flow process. The introduction of particles increases resin viscosity, which in turn influences flow behaviour. Moreover, particle filtration by the fibrous medium changes its permeability, thereby impacting both flow dynamics and particle distribution. The materials used in the infusion process are experimentally characterised, and the resulting parameters served as inputs for the LRI process simulations. Constitutive behavior laws are implemented within the simulation tool. Simulations are then conducted using all characterized inputs and models for validation purposes. These validated models are subsequently employed to assess the infusion process performance.

Abstract: This paper presents a comprehensive fully integrated polymer composite thermoforming process simulation chain developed in ANSYS LS-DYNA®, covering the full manufacturing sequence from automated preform creation to final part cooling. The entire simulation consists of three distinct phases, namely, thermoforming, cooling (within the tool) and manufacturing-induced dimensional distortion after demolding (spring-back), where distortions develop as the part is removed from the tooling and finally cools to room temperature. The simulation framework employs a modular model structure consisting of tooling, a preform holding system, and a detailed preform representation based on a semi-discrete unit cell approach. Individual laminate plies are modeled using a combination of beam, solid, and shell elements to accurately capture temperature-dependent bending, shear, and thermal behavior of the preform. To ensure industrial applicability, an automated preform meshing strategy has also been developed, utilizing tape placement path planning data exported from the automated tape laying process to generate simulation models with minimal manual effort. The simulation results enable the prediction of spring-angle distortions (spring-in or spring-back) and can be validated against experimental distortion measurement data from manufacturing trials of several different representative CFRTP components. The presented approach demonstrates the capability of the newly developed simulation chain to support thermoforming process development, tool geometry compensation, and robust manufacturing of complex thermoformed CFRTP structures.